Utilizing fully homomorphic encryption to implement secure medical computation in smart cities

Abstract

As healthcare is one of major socioeconomic problems in cities, mobile healthcare network becomes one of core components of smart cities, which would improve urban healthcare environment. However, there are wide privacy concerns as personal health information is outsourced to untrusted cloud servers. It is a promising method to encrypt the health data before outsourcing, but how to do diagnosis computations on the encrypted health data remains an important challenge. In this paper, we propose a general architecture of the mobile healthcare network, and define three typical secure medical computations, which include the average heart rate, the long QT syndrome detection, and the chi-square tests. To achieve computations on the ciphertext, we leverage fully homomorphic encryption (FHE) to encrypt the health data. Different from previous related works, we use more efficient Dowlin’s FHE scheme to implement above three medical computations. In our implementation of the average heart rate, only one ciphertext is sent back to the receiver, so homomorphic decryption is needed once. We take an efficient l-bits comparator to implement the long QT syndrome detection, which only needs l XOR operations and one homomorphic multiplication. We first implement the chi-square tests by homomorphic additions and homomorphic multiplications, which can be used to study whether varicose veins is relevant to overweight. Extensive simulations and analytical results show the scalability and efficiency of our proposed scheme.